Server Operational Status Recognition Based on Non-intrusive Energy Monitoring
Qianhao Zhou, Xiaodong Wang, Ruiqiang Huang, Shenhao Zhao, Min Luo, Lianjun Zhang, Shiquan Fan · 2023
As an important carrier for processing data, the stable operation of servers is related to the business continuity, productivity, data security, customer satisfaction, and the availability and convenience of personal services for enterprises. Aiming at the disadvantages of existing server operation status recognition algorithms such as poor practicality and complicated process of manual feature extraction, this paper proposes a method of using non-intrusive devices to collect the electrical energy consumption when the server is operating and combining with deep learning algorithms to achieve operation status recognition. The corresponding electrical energy consumption data are collected under different server operating status, and a one-dimensional convolutional neural network is designed to process and recognize the data with small computational volume and 100% accuracy.